KNIME Analytics Platform vs. TensorFlow

KNIME Analytics Platform has vastly improved our effectiveness when working with large data sets. The self documenting GUI allows analysts to focus on what they are trying to accomplish, not complex code syntax. If we were to use traditional tools, like SQL, work would take much longer and it would be more difficult to collaborate both internally and with clients. Since KNIME Analytics Platform is database oriented, some spreadsheet functions are not supported, which is as it should be. For small data sets we often use Excel vlookup and pivot tables in place of KNIME Analytics Platform. If VBA code is requried, we go to KNIME Analytics Platform as we find VBA to be unstable in Excel.

1. Whenever the problem has the demand for a neural networks based solution, Tensorflow (TF) is a great fit.
2. The tf.dataset API makes it really simple to create complex data pipelines in a few lines of code.
3. tf.estimators API abstracts all the complex computation graph creation logic making it very simple to get started.
4. Eager execution makes it simple to develop a TF graph as debugging the code would be like any other imperative Python program.
5. TF abstracts all the complexities of scaling it to multiple machines. It has various code and data distribution algorithms ready to use.
6. Projects like TensorBoard make monitoring the training process really easy. It also gives the ability to view embeddings without any extra code. Their What-If is extremely useful for poking and understanding a black box model. It also has tools to visualize data to quickly check for anomalies.
7. TF Autograph aims to covert any normal Python code into a distributed program which is quite handy to scale an existing code base.

Incentivized

• Visual programming as oppose to scripting encourages data analysts to reap deeper insights from their data
• Large community contribution in extending the KNIME Analytics Platform into other areas of analytics, e.g. Text Analytics, Predictive Analytics, ML, etc.
• Open source with periodic updates ensures it is equipped to deal with the most sophisticated data analytics use case

• Data pipeline implementation is quite good, loading large amounts of data and pre-process it in an efficient way is no more issue for us
• It supports all major DL algorithms and network layouts such as ConvNets, RNN, LSTMs, Word2Vec, and even the latest transformer architecture
• The abstraction for the device is perfectly done and its support seamlessly for multiple GPU and even TPU will bring a lot of performance gain for enterprise scoped solution while still keep the flexibility
• The TensorBoard is amazing. I haven't seen a similar thing in other frameworks on the market. It allows us to quickly understand and debug the model with the info visualization which makes understanding much better
• A very supportive community, which is the key for sharing the ideas and find the quick and best solutions

Incentivized

• Automation - e.g. RapidMiner Studio provides a Turbo Prep function, where one can get to working on models more quickly (RapidMiner is not open source though)
• KNIME does not replace a regular reporting tool - it is not meant to. However, if I have already spent some time developing a data acquisition and analytical model, it would be nice to be able to deploy, for example, a monitoring or reporting module that would process data autonomously and react accordingly.

Incentivized

• It would be much better if they could provide good documentation and easy ways to understand concepts.
• It is difficult to understand the concept behind for example, Tensor Graph, which takes a lot of time.
• As you have to write everything, it is time consuming to write the implementation of whole neural network. It would be better if they can provide some wrapper library to make things easier.

Incentivized

We are happy with Knime product and their support. Knime AP is versatile product and even can execute Python scripts if needed. It also supports R execution as well; however, it is not being used at our end

The training KNIME Analytics Platform provide helps you get to grips with a product that is already very intuitive. There is a KNIME Analytics Platform way of thinking about addressing problems, but once you understand a couple of patterns which you see again and again in your workflow it all makes sense.

Incentivized

Support of multiple components and ease of development.

Incentivized

KNIME's HQ is in Europe, which makes it hard for US companies to get customer service in time and on time. Their customer service also takes on average 1 to 2 weeks to follow up with your request. KNIME's documentation is also helpful but it does not provide you all the answers you need some of the time.

Incentivized

Community support for TensorFlow is great. There's a huge community that truly loves the platform and there are many examples of development in TensorFlow. Often, when a new good technique is published, there will be a TensorFlow implementation not long after. This makes it quick to ally the latest techniques from academia straight to production-grade systems. Tooling around TensorFlow is also good. TensorBoard has been such a useful tool, I can't imagine how hard it would be to debug a deep neural network gone wrong without TensorBoard.

Incentivized

KNIME Analytics Platform is easy to install on any Windows, Mac or Linux machine. The KNIME Server product that is currently being replaced by the KNIME Business Hub comes as multiple layers of software and it took us some time to set up the system right for stability. This was made harder by KNIME staff's deeper expertise in setting up the Server in Linux rather than Windows environment. The KNIME Business Hub promises to have a simpler architecture, although currently there is no visibility of a Windows version of the product.

Incentivized

Use of cloud for better execution power is recommended.

Incentivized

There are two aspects which put KNIME Analytics Platform ahead of other products. Firstly the fact that KNIME Analytics Platform comes at no cost and no restrictions on its use is an instant winner for any organisation wanting to democratise their data. It means that a client is free to install it on as many machines as they wish without worrying about costs, the number of seats required or payment models or procurement negotiation. It also means that we are not building costs into our clients business. Secondly, KNIME Analytics Platform has a very comprehensive set of tools for importing/exporting data, data manipulation and data science. Some products offer analytics packages on top of their base offering at additional cost and they are still not as comprehensive as what you get with KNIME Analytics Platform for free. For some types of analysis you may require to download additional packages with KNIME Analytics Platform, but its invariably at no cost, those packages are kept out of the main download to keep the size down. Due to the easy integration with R and Python, I view KNIME Analytics Platform as also having the capabilities of those languages too. This has helped me in the past with seamlessly importing a rare filetype and using very specific models not directly available in KNIME Analytics Platform.

Incentivized

Can't seem to choose any deep learning platform in the above, so I'll list it here: 1. Apache MXNet: this has been used for one of our main algorithms for search as an end-to-end pipeline. We chose this because of the Scala bindings, which makes it easier to integrate with out JVM backend. MXNet seems comparable to TensorFlow, although community support is not as good as TensorFlow, and there are issues with memory leaks that are being worked on. TensorFlow in general is easier to use, but MXNet isn't too far behind. 2. Keras: still a favorite. Often I use this when paired with TensorFlow. TensorFlow 2.0 will make it even easier. 3. PyTorch: only used it a little, so it's hard to provide a good opinion. 4. DL4J: used it initially in an early days project because it has good JVM support. Harder to used not because of poor API design, but because community support is lacking and features don't come out as fast as TensorFlow.

Incentivized

• It is suited for data mining or machine learning work but If we're looking for advanced stat methods such as mixed effects linear/logistics models, that needs to be run through an R node.
• Thinking of our peers with an advanced visualization techniques requirement, it is a lagging product.

Incentivized

• Positive Impact- As I mentioned before its open source. Very easy to learn for average programmer/ developer. We were able to design a POC model for understanding the patient appointment cancellation snd reasons behind it in 3 week time frame.
• Negative Impact- If you are using tensor flow for small project it works fine. If you are trying to build a model for face recognition it will be hard to program and train the system. It needs data to be processed before hand cannot learn on the go.